ISO standardized mp3's core coding format in August 1993

The audio task

An engineer choosing a bitrate, or a distributor setting a delivery spec, is working inside audible-fidelity expectations that were substantially set thirty years ago by a specific standardization decision, one that still frames how listeners and reviewers judge whether a compressed or generated track sounds real.

What the documents show

ISO's own standard catalogue record states that ISO/IEC 11172-3:1993, the audio part of the MPEG-1 standard, was published on 12 August 1993, specifying the coded representation of high-quality audio for digital storage media at roughly 1.5 megabits per second. Fraunhofer IIS's own mp3 page states that development of what became mp3 started in the late 1980s at Fraunhofer IIS, building on earlier research at the University of Erlangen-Nuremberg, and that its mp3 patent licensing program with Technicolor was terminated on 23 April 2017. Fraunhofer's own retrospective article, published July 4, 2025, adds that the ISO-MPEG committee standardized the format across three audio layers, that the specific name mp3, short for MPEG-1 Audio Layer-3, was chosen on 14 July 1995 through an internal email poll at the institute, and that the format works by discarding audio components the human ear cannot perceive, shrinking files up to tenfold with, the article states, virtually no audible difference from the source.

Rights status

These are technical and corporate-history records, not rights determinations. Fraunhofer's own page is explicit that it does not sell mp3 products to end users or provide end-user support, and that its patent licensing program for mp3 specifically has already ended. None of the three documents addresses whether a given AI tool's use of mp3-format training data or output raises a separate rights question; they describe a codec's origin and patent lifecycle, not content licensing.

What to check before you use it

This is an editorial checklist. Before assuming an mp3-related patent claim still applies to a modern tool, confirm the 2017 licensing-program termination against the specific patent family in question. Check whether a fidelity comparison to mp3 quality is being used loosely as shorthand or as a specific, testable bitrate claim. And verify a vendor's marketing use of lossless or CD quality against the same perceptual-coding logic Fraunhofer describes, rather than assuming the terms are interchangeable.

• Does a cited mp3-patent claim predate or postdate the 2017 licensing-program termination?
• Is an audio-quality comparison naming a specific bitrate, or only an unverified impression?
• Does a lossless marketing claim describe the same perceptual-coding trade-off mp3 relies on, or a different one?

mp3's standardization is a codec-design and corporate-licensing story with a precise paper trail. It is a useful baseline for fidelity debates about generated audio, but it answers questions about compression, not about who may license or train on the recordings compressed.